The present invention relates generally to image processing of a camera, and more particularly to combination of a voice command and image recognition to provide a mechanism for identifying a subject of interest within a photo or a video image frame.
When using a camera, often it is frustrating that the camera does not recognize what is a subject of interest in a photo or video frame. This situation may cause issues when attempting to set parameters such as focusing, zoom, and exposure. Also, where cameras have the ability to rotate and pan, keeping a subject of interest in the frame can be difficult.
With facial recognition, a camera can locate a face within a photo or video image frame and then make the face a subject of interest. To a certain extent, the facial recognition works with a face within the frame but does not work for a subject other than a face or multiple faces in the frame. Face detection provides a means to identify faces in image frames. This is valuable so long as a user of a camera is interested in a face or faces identified by the camera. If a subject of interest is not a face or there are multiple faces and the user of the camera is interested in one in particular, then facial recognition does not serve to identify a subject of interest.
Websites such as Facebook and Google have technology of facial identification in photos to identify faces within a photo. Although facial identification can be done for a video or photo image, this technique itself does not provide a mechanism for identifying the subject of interest while using a camera.
It can be possible to use a touch screen to identify a subject of interest in a photo or video image frame. However, the screen sizes on many devices and the difficulty of viewing in the bright sunlight can make it tricky to identify a subject of interest. Additionally, with touch screens, it is difficult to select a distant subject when it is partially occluded by a closer subject.